Hiking Intensity Along the Arizona Trail Steven Woodruff - CEE 187 - Tufts University -15 December 2015

Introduction Before going on any hike, the first thing one must do is pre-

pare. Collecting data on any hiking trail presents the challenge

of trying to pull and combine all the important parameters that

go into a successful hike. Information about the difficulties a

trail will present are often subjective - derived from the per-

sonal accounts of people who hiked the trail before. Pulling

information from stories and tips is cumbersome and ineffi-

cient and the variety of data about the trail can be incompre-

hensible. The purpose of this project is to combine the most

important parameters that determine how difficult or danger-

ous a trail will probably be to create a map showing the inten-

sity of each portion of the trail to allow the people who are

preparing to hike to have a convenient, comprehensive over-

view of the trail.

The scope of this project is limited to analyzing the Arizona

Trail (AT), a National Scenic Trail. This 800+ [mi]

trail extends across the state of Arizona from its border with

Mexico north to the border with Utah. Crossing deserts and

mountains, the AT requires careful planning. Besides rule-of-

thumb time estimates, there are no commonly used models to

predict the difficulty of the AT at different sections. This

model displays the difficulty of each point on the trail relative

to the entire system.

With this map, hikers can make better-informed decisions on

how to prepare for this 30+ day hike in order to prevent injury

on the trail. This map could also be used to better predict

where a hiker was injured so that rescue forces could more ef-

ficiently locate the victim.

1. Georeference trail image raster—make trail polyline

2. Reduce US-wide data to intersecting points and relevant regions using spatial and attribute queries

3. Convert trail polyline to raster of value 1

4. Use Path Distance tool (Spatial Analyst) to create distance rasters from intersecting points & urban center polygons

5. Multiply distance rasters & DEM by the trail raster to get trail distances & elevations —model values

6. Use Spatial Analyst to find slope & aspect (to determine sign of slope South to North) along trail elevation raster

7. Normalize & weigh data parameters per model

8. Add rasters to get model raster map

9. Convert raster model into polyline

Relative Intensity Model:

I = 0.1Dr + 0.25Dw + 0.2Du + 0.3PS + 0.15NS

where: I = relative intensity

Dr = normalized distance from last road intersection

Dw = normalized distance from last river intersections

Du - normalized distance from last urban center

PS - normalized positive slope (uphill)

NS - normalized negative slope (downhill)

Data 1. Image raster of trail

2. Digital Elevation Model (DEM) of North America

3. Vector data for: US urban centers, US rivers & stream, US

roads

Methodology

Results 1. Individual parameter distributions are con-

tinuous and segmented

2. Model intensity is normally distributed

3. Model average value is ~25%

4. Differences in parameter values at same

points can be large

5. Model weights greatly affects outcome

6. Model lacks true objective reasoning

Conclusion When planning to leave the comforts of civi-

lized life for the backcountry, knowing where

the most dangerous or difficult sections of the

trail are can be life-saving. Knowing that a

long stretch of red or orange on the map is

coming up tells the hiker that to carefully pre-

pare. With more information and a more sci-

entifically refined model, this model map

could save lives and increase the pleasure of

long-distance hiking.

References 1. Price, Maribeth. Mastering ArcGIS. McGraw-Hill. 7th Edition. 2015.

2. ESRI Data & Maps 10

3. The Arizona Trail Association. 2015. Website.